Saw this on YouTube. Left this comment elsewhere but thought it might be useful here:
I really appreciate the “anti-debate” format of this and look forward to more. It was really interesting to see these two work towards something like synthesis/synergy in real time. It’s a little messy, but that it’s good to see that because that’s how it is in the real world too.
That said, there were times when I wished the guests could have prepared more for the sake of clarity. There were a few points I wasn’t clear about until I threw the transcript into Claude along with searching Daniel and Dean’s publicly available work to fully understand where one or the other was coming from.
Happy to see that they both support mechanisms that would force the labs to be transparent about their model specifications and values, allowing the rest of society (judiciary, public, and auditors) time to react to the rapid pace of change.
This was a meaningful contribution to better sense making on our uncertain AI future. Thanks Liv and Stephanie!
Thanks for the feedback Glen! Yes I think this has a lot of potential but as you say it very much depends on the participants (and my!) levels of pre-prep and also ability to track each other’s points, which on this first go I found difficult. I imagine it’ll get easier the more we do though.
Saw this on YouTube. Left this comment elsewhere but thought it might be useful here:
I really appreciate the “anti-debate” format of this and look forward to more. It was really interesting to see these two work towards something like synthesis/synergy in real time. It’s a little messy, but that it’s good to see that because that’s how it is in the real world too.
That said, there were times when I wished the guests could have prepared more for the sake of clarity. There were a few points I wasn’t clear about until I threw the transcript into Claude along with searching Daniel and Dean’s publicly available work to fully understand where one or the other was coming from.
Happy to see that they both support mechanisms that would force the labs to be transparent about their model specifications and values, allowing the rest of society (judiciary, public, and auditors) time to react to the rapid pace of change.
This was a meaningful contribution to better sense making on our uncertain AI future. Thanks Liv and Stephanie!
Thanks for the feedback Glen! Yes I think this has a lot of potential but as you say it very much depends on the participants (and my!) levels of pre-prep and also ability to track each other’s points, which on this first go I found difficult. I imagine it’ll get easier the more we do though.
Excellent format, guests, and content. Compelling insights near the 1-hour mark. The most important topic at the moment.
Thanks Craig! That means a lot. Feel free to share it if you feel inclined, I’m trying to get it out there.
"AI" is just software.
There is no where in the Constitution the right for our servants in government to control what we say, or write, or draw, or code....
indeed! But perhaps watch the conversation to see why it is not just software...
Where might I send a thought piece for your consideration?
That includes historical context; mankind's move from physical to intellectual adaptation; and divergence of Dry Life from Wet.
Thanks for putting this kind of stuff out there!